Re: [Boost-testing] -pedantic and -Wno-long-long

Subject: Re: [Boost-testing] -pedantic and -Wno-long-long
From: Vicente Botet (vicente.botet_at_[hidden])
Date: 2011-01-30 05:48:08

John Maddock-3 wrote:
> 
>>> >Some testers use to add -pedantic to the cxxflags on GCC based
>>> compilers 
>>> >as
>>>>gcc, pathscale and darwin.
>>>>
>>>>As a lot of libraries use long long inside his code, shouldn't these
>>>>testers add also -Wno-lon-long ?
>>>
>>> At least in theory as long as you're using boost::long_long_type and
>>> boost::ulong_long_type in from config.hpp then this shouldn't cause you
>>> a
>>> problem.
>>
>> Oh, I remember this. The issue was that boost::intmax_t and 
>> boost::uintmax_t don't use this workaround on GCC. Can this be changed, 
>> that is, can boost::intmax_t uses boost::long_long_type when
>> std::intmax_t 
>> is a typedef for long long?
> 
> Ah.... actually they do use that workaround when we define those types 
> ourselves... but on most platforms we simply include the platform's
> stdint.h 
> so we have no real control over those definitions :-(
> 
> So the only suggestion I have is to add a 
> <toolset>gcc:<cxxflags>-Wno-long-long to your usage requirements.
> 
> 

Yes. this is what I did. But need to add also all the toolsets that are
based on a gcc compiler as darwin, pathscale and others I don't know yet. Is
for this reason I suggested the tester to add this flag every time them add
-pedantic.

Vicente

-- 
View this message in context: http://boost.2283326.n4.nabble.com/pedantic-and-Wno-long-long-tp3236832p3247025.html
Sent from the Boost - Testing mailing list archive at Nabble.com.